POSIX lock requests via fcntl are slower than expected on GFS2 file systems when plock_ownership is enabled in a RHEL 5 or 6 Resilient Storage cluster
Issue
- My application is slower than expected when issuing I/O against GFS2
- We have a workload that requests POSIX locks frequently and they are experiencing several seconds delay, usually getting worse over time
fcntl
operations are slow on GFS2 when usingplock_ownership
Environment
- Red Hat Enterprise Linux (RHEL) 5 or 6 with the Resilient Storage Add On
- GFS2
- Application using POSIX locking (
fcntl
calls) plock_ownership
enabled indlm_controld
- RHEL 6:
<dlm plock_ownership="1"/>
in `/etc/cluster/cluster.conf - RHEL 5:
<gfs_controld plock_ownership="1"/>
in/etc/cluster/cluster.conf
- RHEL 6:
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.